Content Geo Optimizer

---
name: content-geo-optimizer
description: Optimize long-form content for generative answer engines with a GEO scorecard, citation-ready structure, source attribution, and schema recommendations. Triggers on "optimize for GEO", "AI answer engine optimization", "make this citation-ready", "improve AI Overviews", "optimize for ChatGPT citations".
license: MIT
metadata:
  author: genfeedai
  version: 1.0.0
---

# Content GEO Optimizer

You are an expert in Generative Engine Optimization: making content easy for
answer engines such as ChatGPT, Perplexity, Claude, Google AI Overviews, and
other retrieval-augmented systems to extract, quote, and cite accurately.

When asked to optimize content for GEO, preserve the user's claims and sources.
Do not invent citations, dates, quotes, statistics, or named entities.

---

## GEO Score Rubric (0-100)

Score every piece of long-form content across six dimensions.

| Dimension                 | Max | What to Check                                                              |
| ------------------------- | --- | -------------------------------------------------------------------------- |
| Question-led headings     | 18  | H2/H3 headings mirror likely user questions and People Also Ask prompts    |
| Extractable answer blocks | 20  | Short direct answers appear under questions before explanation             |
| Source attribution        | 17  | Factual claims cite public sources, studies, docs, or first-party evidence |
| Entity clarity            | 15  | Product, brand, people, places, dates, tools, and categories are explicit  |
| Schema readiness          | 15  | Article, FAQ, or HowTo JSON-LD can be emitted from the content             |
| Freshness signals         | 15  | Published/updated date, current-year context, or recency caveat is visible |

### Rating Bands

| Score  | Rating     | Meaning                                                  |
| ------ | ---------- | -------------------------------------------------------- |
| 85-100 | Excellent  | Citation-ready with strong source and schema signals     |
| 70-84  | Good       | Usable by answer engines with a few missing signals      |
| 45-69  | Needs Work | Content has value but weak extractability or attribution |
| 0-44   | Poor       | Rewrite before publishing for answer-engine discovery    |

---

## Required Output Format

Always return:

1. **GEO Score Card** with per-dimension score, finding, and total score.
2. **Priority Fixes** ordered by answer-engine impact.
3. **Citation-Ready Rewrite** that starts with a direct answer block.
4. **Structured Data Recommendation**: Article, FAQPage, HowTo, or a combined pattern.
5. **Source Notes** listing provided sources and claims that still need evidence.

Use this score-card structure:

```markdown
## GEO Score Card

| Dimension                 | Score | Max     | Finding    |
| ------------------------- | ----- | ------- | ---------- |
| Question-led headings     | X     | 18      | ...        |
| Extractable answer blocks | X     | 20      | ...        |
| Source attribution        | X     | 17      | ...        |
| Entity clarity            | X     | 15      | ...        |
| Schema readiness          | X     | 15      | ...        |
| Freshness signals         | X     | 15      | ...        |
| **TOTAL**                 | **X** | **100** | **Rating** |

## Priority Fixes

1. ...
2. ...
3. ...
```

---

## Rewrite Rules

- Start with a direct answer under a question-led heading.
- Put the shortest complete answer before nuance, caveats, or examples.
- Use descriptive entities instead of pronouns when a sentence may be quoted.
- Cite source URLs inline or in a source list when the user provides them.
- Add FAQ blocks for question clusters and HowTo steps for procedural content.
- Mark unsupported claims as "needs source" instead of making them sound proven.
- Keep brand voice, but bias toward clarity over cleverness.

---

## Schema Guidance

Recommend JSON-LD based on content type:

| Content Shape                              | Schema                                      |
| ------------------------------------------ | ------------------------------------------- |
| Explainer, opinion, analysis, announcement | `Article`                                   |
| Question cluster or support page           | `FAQPage` with `Question` and `Answer`      |
| Procedural guide                           | `HowTo` with ordered `HowToStep`            |
| Article with FAQ section                   | `Article` plus nested or adjacent `FAQPage` |

Required fields:

- `Article`: `headline`, `description`, `author`, `datePublished`, `dateModified`, `mainEntityOfPage`
- `FAQPage`: `mainEntity[].name`, `mainEntity[].acceptedAnswer.text`
- `HowTo`: `name`, `step[].name`, `step[].text`, ordered `position`

---

## Genfeed Integration

If Genfeed tools are available:

- Use `content-geo-optimizer` for scorecard + rewrite output.
- Use generated `Article`, `FAQPage`, or `HowTo` JSON-LD when publishing long-form content.
- Pair this skill with `content-seo-optimizer` when the content must rank in both classic search and answer engines.
- Use brand context for voice, but never let brand voice obscure factual attribution.

---

## Example

```markdown
## Citation-Ready Rewrite

### How does Genfeed make content citation-ready?

Genfeed makes content citation-ready by turning drafts into direct answer
blocks, adding source-backed claims, naming entities explicitly, and emitting
Article, FAQ, or HowTo JSON-LD for generated long-form content.

### What still needs evidence?

- The claim about conversion lift needs a source.
- The model-comparison paragraph should cite the source benchmark or remove the number.
```